The “Bird Plot”

Jessica Esquivel gave a very nice update last week at the MicroBooNE collaboration meeting about her work to use “deep learning” tools to separate muons and pions in the CC-Inclusive cross-section analysis.  She included this animation, which shows a “deep learning” algorithm learning how to distinguish between five different particle types.  This work was done using the SUrge GPU cluster at Syracuse University.

t-distributed stochastic neighbor embedding algorithm learning features to distinguish among five particle types.